THE GENDER DYNAMICS OF INCLUSION AND EXCLUSION: A MACRO PERSPECTIVE
D. Inclusion and exclusion in employment: Gender trends 9
conflictive upper left quadrant, with nearly two thirds witnessing declines in men’s employment. Some of these declines were quite significant (e.g. more than 5 percentage points in Kenya, Mauritius, Nigeria and South Africa). The vast majority of countries in the developing America region (77.3 per cent) are in the upper left quadrant, with increases in women’s relative employment as men’s employment declined;
the other countries from that region are in the upper right quadrant, showing increases in both women’s and men’s employment rates.
While women’s employment has been rising in most countries (with some notable exceptions) regardless of level of development, the associated improve-ment in gender equality – as measured by women’s employment relative to men’s – has been partly driven by substantial declines in men’s employment.
Given the push and pull factors driving women’s labour force participation, highlighted above, it is worrying that distress sales of labour might be play-ing a role in what superficially appears to be greater gender equality in employment. That is, women’s higher relative employment rates in a number of countries are likely to be due not to job competition between women and men, but rather, to women taking on inferior jobs in order to maintain family incomes in response to men’s declining job opportunities and slow wage growth. This highlights the importance of achieving “inclusive” gender equality, in the sense of improvements for women not being at the expense of men. This partly depends on the overall state of an economy. Increasing women’s employment partici-pation without addressing demand-side constraints, or acknowledging the widespread failure of growth – when it occurs – to generate good jobs, will merely escalate labour market competition, ultimately to the detriment of both women and men.
2. Industry and “good” jobs
As noted above, gender stratification plays an impor-tant role in allocating jobs within segmented labour markets, especially as competition for core sector work intensifies. Although women’s employment relative to men’s has been rising in most developing countries for more than two decades, their share of
“good” jobs has been falling. That is, during the past 25 years of growing global integration, women have been increasingly excluded, as compared to men, from prized jobs, even as their educational attain-ments and labour force participation have risen. In this chapter, jobs in the industrial sector (rather than agricultural or services sectors) are used as a proxy for “good” jobs, for reasons outlined below.
In most trajectories of productivity-enhancing structural change and development, the processes of industrialization and the shifting of resources – including labour – into higher productivity sectors support aggregate productivity growth. However, it is through the expansion of higher productivity work in the modernizing, increasingly diversified industrial sector that labour initially accesses the higher incomes
FIGURE 4.2 Changes in women’s to men’s employment rates versus men’s employment rates, 1991−2014 (Percentage points)
Source: UNCTAD secretariat calculations, based on ILO modelled employment rates.
Note: Employment rates refer to the proportion of the wage-earning population, aged 15 years and older. Changes are percentage point changes in 3-year average values. Also note that the axes in figure B are different from those used in figure A in order to better illustrate regional differences.
-10 0 10 20 30 40
-20 -10 0 10 20
Change in women’s to men’s employment rate
Change in men’s employment rate Developed countries Developing countries Transition economies
-10 0 10 20 30
-15 -10 -5 0 5 10 15
Change in women’s to men’s employment rate
Change in men’s employment rate Africa America Asia
A. By developed, developing and transition economies
B. By developing region
THE GENDER DYNAMICS OF INCLUSION AND EXCLUSION: A MACRO PERSPECTIVE ON EMPLOYMENT
that accompany industrialization and development, ultimately building domestic aggregate demand and sustaining aggregate productivity growth. (In this sense, for growth to be sustained it must also be inclusive.) When these connections fail to materialize, or weaken, stalled or premature (de)industrialization dampens the prospects for inclusive development.
Higher value-added, knowledge-intensive services, which account for a more substantial share of employ-ment than industry in developed countries, have recently been emphasized as an alternative to the lacklustre job-generating performance of industry in developing countries. However, in developing countries, in particular, the services sector alone is not likely to provide a sufficient alternative to indus-try for the generation of “good” jobs, especially if it is disconnected from a dynamic industrial sector (Kucera and Roncolato, 2016; TDR 2016). Relative to the industrial sector, jobs in the services sector are more likely to be informal and insecure, with lower productivity and thus lower wages, especially for women. They most probably reflect the growth of low-productivity (often traditional) services rather than the beginnings of long-term dynamism − a type of disguised unemployment that ultimately reflects the failure of growth to generate enough decent work. Accounts of the links between globaliza-tion and informalizaglobaliza-tion echo these problematic dynamics (Bacchetta et al., 2009). Even India, which is often cited as an exemplar of the growth of high-productivity services as a conduit for growth and development, has failed to produce many good jobs in this sector (Chandrasekhar and Ghosh, 2014).11 Measures of decent work, as defined by the Inter-national Labour Organization (ILO), provide a good basis for comparing the quality of employment in services and industry. Decent work is defined as work that is productive, has workplace protections, and offers social protection and prospects for indi-vidual development (such as skills upgrading). In the absence of an international dataset on decent work opportunities by sector, a measure of relative job quality can be calculated using the ratio of labour pro-ductivity in the services sector to that in the industrial sector (see table 4.1 by region). The rationale for this comparison is that higher productivity measures are associated with greater remuneration and benefits.
This is not the same as saying that industrial workers are more “productive” than services sector work-ers. Trying to measure services sector productivity is controversial, partly because of the difficulty in
measuring outputs. Indeed, for the services sector at least, productivity measures can be thought of more as a consequence of wages than a cause. Hence, higher relative productivity in developed countries in this sector partly reflects higher per capita incomes.
Regardless, lower productivity measures indicate lower wages. Among developing regions, to varying degrees, services sector labour productivity is lower than industrial labour productivity (with ratios less than 1). The median for all non-developed regions is close to 0.75, suggesting that average productivity is roughly 25 per cent lower in the services sector than the industrial sector.
Based on these data, for developing countries, there is a positive association between the services sector’s relative productivity and the relative concentration of men in that sector. That is, the higher the aggregate labour productivity in the services sector relative to the industrial sector, the higher too is men’s con-centration in that sector relative to women’s (with a correlation of 0.43 for the developing countries in the sample). To the extent that these measures of relative productivity mirror relative wages, this outcome is in line with the predictions about how gender stratifica-tion manifests in dual or segmented labour markets:
the better the jobs, the more likely it is that members of the dominant group will “opportunity hoard”, and thus the less likely it is that members of the subordi-nate group, in this case women, will have those jobs.
Given that jobs in the industrial sector are more likely to be part of the core labour market (that is, formal jobs with associated benefits and protections) than jobs in the agricultural or services sectors, this chapter
TABLE 4.1 Ratio of services sector to industrial sector labour productivity, 1991–2015
Region Mean Median
Full sample 0.89 0.87
Developed countries 1.04 1.05
Developing countries 0.79 0.75
Africa 0.83 0.75
America 0.72 0.74
Asia 0.82 0.74
Transition economies 0.83 0.75
Source: UNCTAD secretariat calculations using the World Bank, World Development Indicators (WDI) and Penn World Tables databases.
Note: Sectoral productivity is calculated as the value added of sectoral output relative to the number of employees in that sector;
unweighted means and medians for country groups are for the period 1991−2015.
uses relative access to industrial jobs as a proxy for gender equality.12 Evaluating the absolute and rela-tive trajectories of employment in this sector affords insights into whether and to what extent growth has been inclusive from a gender perspective.
3. Women’s exclusion from “good” jobs At the outset, it is important to note that the overall availability of industrial sector jobs has declined since the early 1990s, for both women and men. On average, industrial sector employment as a percent-age of total employment declined in all groups of countries − developed, developing and transition economies (figure 4.3). The decline was the most pronounced in developed countries. Using three-year averages to compute changes in the share of indus-trial employment in total employment, developed and developing countries and transition economies experienced declines of -7.8, -3.5 and -5.2 percentage points, respectively, between 1990 and 2014.
The kernel density functions displayed in figure 4.4 provide evidence of the degree of sectoral employ-ment segregation by gender in developing countries in 2013. This figure shows the distribution of coun-tries according to two ratios that compare women to men: women’s employment-to-population rate rela-tive to men’s, with a sample mean of 61.8 per cent;
and the ratio of women’s concentration in industrial employment to men’s concentration, with a sample
mean of 47.2 per cent. The latter measure is referred to as “women’s relative concentration in industrial employment” for the remainder of the chapter, and it proxies for women’s relative access to good jobs. As illustrated by the shapes of the curves in figure 4.4, women’s relative concentration in industry is much more widely dispersed, and lower, on average, than women’s relative employment participation overall.
This is evidenced by a decline in women’s relative employment concentration in the industrial sector since 1991, from an average of 70.2 per cent in 1991 to 47.2 per cent in 2013 (figure 4.5).13 This phenom-enon occurred in all developing-country regions, with African countries showing the largest decline (table 4.2). Even in Asia, where industrialization and export-oriented manufacturing have been more substantial, a decline in women’s concentration in
“good” jobs in the industrial sector can be observed, although their relative share in employment rose.
Figure 4.6 contrasts the distribution of developing countries by percentage point changes between 1991 and 2013 for two measures of women’s relative (to men’s) employment share: in total employment and in the industrial sector. The horizontal axis displays
FIGURE 4.3 Trends in industrial employment as a share of total employment, 1990−2014
Source: See figure 4.2.
Note: Values refer to the unweighted average by year for country group, which is consistent across years.
FIGURE 4.4 Distribution of developing countries by women’s to men’s economy-wide employment rates and shares of industrial sector jobs, 2013
Source: See figure 4.2.
Note: Women’s relative concentration is calculated as three-year average of the share of women employed in the industrial sector relative to men’s share. Developing country group is consistent across figures 4.4, 4.5 and 4.6, and differs from the (larger) group illustrated in figure 4.1, as the current group is limited to countries for which there is data on women’s industrial share of employment across the particular years considered.
15 20 25 30 35
1990 1994 1998 2002 2006 2010 2014
Developed countries
Developing countries Transition economies
.03
.02
.01
0
mean = 61.8
mean = 47.2
Density
Ratio of women to men
Employment to population (economy−wide) Concentration in industrial employment
0 20 40 60 80 100
THE GENDER DYNAMICS OF INCLUSION AND EXCLUSION: A MACRO PERSPECTIVE ON EMPLOYMENT
values for the percentage point changes, while the vertical axis gives the corresponding incidence or percentage density of each of these values. What is important to consider is the shape of each curve, and its relative position along the horizontal axis.
The ratio of women’s to men’s total employment increased, on average, by 9.2 percentage points, and countries are tightly grouped around that average, as illustrated by the steep curve. The curve is also centred primarily on positive values, as illustrated by its position relative to the vertical zero-intercept
line, indicating an increase in women’s relative total employment in the vast majority of countries (91 per cent) over the period. Conversely, women’s share in industrial employment relative to men’s declined by an average of 23 percentage points, and most of the curve is situated to the left of the zero-intercept line, illustrating that the vast majority of countries (88 per cent) experienced a decline of women’s share in industrial employment relative to men’s.
Figure 4.7 shows the same percentage point changes in women’s relative industrial employment as in figure 4.6, except that here it is by individual coun-try, and is juxtaposed against the percentage point changes in men’s concentration in industrial employ-ment. Women’s relative share in industrial jobs declined in the vast majority of countries between 1991 and 2013; in about half of these cases, men’s share in industrial employment declined as well. This is indicative of both a reduction in labour demand in the industrial sector and of women’s industrial employment rate falling faster than men’s. These patterns provide evidence of gender-based job ration-ing: as industrial sector employment has declined, women’s access to that employment has become more restricted.
Taken together, these figures indicate that over the past 25 years, gender stratification in labour mar-kets has become worse, in the sense that women are increasingly excluded from good jobs, and are,
FIGURE 4.5 Women’s relative concentration in industrial sector employment, developing countries, 1991 and 2013
Source: See figure 4.2.
TABLE 4.2 Employment ratios of women to men, and relative concentration of women in industrial employment, by developing region, 1991 and 2010
(Per cent)
Developing region
Ratio of women’s to men’s employment
rates
Relative concentration of women in industrial
employment
1991 2010 1991 2010
Africa 53.0 57.2 91.8 47.9
America 48.0 61.1 67.9 53.1
Asia 46.3 51.0 59.3 47.2
South Asia 42.0 46.7 63.8 40.8
East Asia 62.2 73.2 75.9 33.1
West Asia 25.2 28.0 22.1 36.5
South-East Asia 62.8 66.9 87.9 66.1 Source: UNCTAD secretariat calculations, based on ILO data, extracted
from the World Bank, WDI database (accessed 15 February 2017).
Note: The data are based on three-year averages.
FIGURE 4.6 Change in women’s relative
concentration in industrial employment and total employment in developing countries, 1991−2013
Source: See figure 4.2.
1991mean = 70.2 2013mean = 47.2
Density
0 50 100 150
Women’s relative concentration in industrial employment .020
.015
.010
0 .005
mean = 9.2
mean = −23.0
Density
−100 −50 0 50
Percentage point change in women’s relative total employment and industrial employment
Women’s relative industrial employment Women’s relative total employment .05
.04
.01 0 .03 .02
instead, crowded into work that is less remunerative and secure. Thus, contradictory forces appear to be at work in developing-country labour markets:
women’s increasing relative share of paid jobs, but their growing exclusion from “good jobs”, suggesting the “crowding” of women in poor quality employ-ment. This process has occurred in the context of the industrial sector’s weakening role as a generator of high-quality employment, manifested as deindustri-alization in developed and middle-income economies and stalled industrialization or premature deindustri-alization in developing countries (TDR 2016).
The decline in women’s relative concentration may also be due to the changing structure of the industrial sector itself, coupled with relatively rigid gender-differentiated employment in that sector. As countries upgrade to more skill- or capital-intensive production and away from labour-intensive production, where women’s employment has been most notable, a fall-ing concentration of women in the industrial sector may result. Indeed, it has been found that in the manufacturing sector, a process of defeminization of employment has been occurring since the mid-1980s (Kucera and Tejani, 2014; Tejani and Milberg, 2016).
FIGURE 4.7 Percentage point changes in women’s relative and men’s absolute concentration in industrial employment, selected economies, 1991−2013
Source: See figure 4.2.
Argentina
Brazil Chile
Colombia
Costa Rica Dominican Republic Ecuador
Egypt El Salvador
Ghana Guatemala
Honduras Hong Kong SAR, China
Indonesia Jamaica
Republic of Korea
Malaysia Mauritius
Mexico
Morocco
Namibia
Pakistan
Panama Paraguay
Philippines Saudi Arabia
Singapore
United Republic of Tanzania
Thailand Turkey
Uruguay
Zambia
-100 -75 -50 -25 0 25
-20 -15 -10 -5 0 5 10
Women’s relative concentration in industrial employment
Men’s concentration in industrial employment